Date: Sun, 14 Jan 2007 02:50:02 +0300 From: "Andrew Pantyukhin" <infofarmer@FreeBSD.org> To: "Doug Barton" <dougb@freebsd.org> Cc: FreeBSD Ports <ports@freebsd.org>, Robert Huff <roberthuff@rcn.com> Subject: Re: Restricting (human) language and character set in /usr/ports Message-ID: <cb5206420701131550n54bb2bcdo52aa04798f9d4ff7@mail.gmail.com> In-Reply-To: <45A96382.3040407@FreeBSD.org> References: <cb5206420701131101l21807993ld15c899e6754ec02@mail.gmail.com> <17833.15710.478810.6251@jerusalem.litteratus.org> <cb5206420701131339s66f2b3f9s2924a55c429ef501@mail.gmail.com> <45A96382.3040407@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On 1/14/07, Doug Barton <dougb@freebsd.org> wrote:
> Andrew Pantyukhin wrote:
> > On 1/13/07, Robert Huff <roberthuff@rcn.com> wrote:
> >>
> >> Andrew Pantyukhin writes:
> >>
> >> > I'm not sure if there's a policy already, but it seems
> >> > we have discussed this before.
> >> >
> >> > Can we limit /usr/ports (the whole ports collection) to
> >> > English language and ASCII characters? This restriction
> >> > should probably apply to all text data (with possible
> >> > exception for patches).
> >>
> >> I don't follow this issue (much), so could you explain what's
> >> broken about the /status quo/?
> >
> > It depends on what you mean by /status quo/, but in
> > short, when I look at COMMENT, pkg-descr, pkg-message,
> > comments in Makefile and other such text data, I
> > expect to see English language and ASCII characters.
> >
> > There are ports that don't follow this expectation and
> > I'd like to change that.
>
> I'm not sure it's quite so cut and dry as that. For example, I think
> it's probably reasonable for the /usr/ports/<language> ports to have
> some non-ascii stuff to start with.
I'm not against non-ascii, but there's no notion
of character set in /usr/ports. I work in UTF-8
and it's visible to me, as it is to many other
people working in many different charsets.
There are several ways to deal with non-ascii
characters, the two most effective being:
(a) select a universal charset (I would love to
see UTF-8 in that role)
(b) introduce special markers, defining current
charset
I can not agree with people selecting random
charsets and me having to guess them. Automated
information brokers, like freshports, also have
problems with this.
As for the language, I expect everything within
the FreeBSD project to be present at least in
English. L10n overlaps with multiple charset
support and without both concepts implemented
one way or another (like they are in FDP), I
think presenting content exclusively in English
is the only solution.
> Is there a problem you're trying to solve here, or is this just a
> matter of tidying things up a bit?
To me it looks like a matter of consistency and
usability.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?cb5206420701131550n54bb2bcdo52aa04798f9d4ff7>